Various abnormal behaviors often accompany fraudsters before and after their fraud activities. These behaviors may occur at different business points. Therefore, we hope that you can transfer these business data through the event API before you call the decision API. In this way, subsequent decisions can be more accurate, for example:
- It helps us identify more cunning fraudsters. These fraudsters are often more challenging to identify. They usually use some standard identities or behaviors to bypass some key business points to deceive the system to achieve the purpose of fraud. A single business behavior judgment is difficult for the machine to know the complete picture of the fraudster. For example: Fraudsters have abnormal behaviors when registering and logging in scenario, but they use other normal identities or behaviors during the transaction process, and then they pretend to be an excellent person to deceive the system into completing the transaction, when the transaction is completed, fraudulent behaviors such as fraud chargeback and malicious return will be carried out. It can be seen that it is hard to identify such camouflage only by the data of a single point. Obtaining the data of the whole picture before the transaction will help us to describe the fraudster's portrait more thoroughly and accurately.
- It will help us more accurately identify who are the good users and improve the conversion.
- Decisions are made at crucial points of the business, and the data transfer happens before that, this way will even not be noticed by the users, helping to reduce disruption to good users and improve their experience.
In different scenarios, the events that need to be transferred are different. Although not all events are required to be transferred, the more events transferred, the more accurate the decision result will be.